Various studies have explored the impact of the keyword method on intentional vocabulary learning (see the critical discussion of the keyword method under Section 2.4.4). However, few studies have addressed the impact of other self-selected strategies on the deliberate learning of L2 vocabulary. Examples of such studies include Lawson and Hogben (1996), and Barcroft (2009), which are discussed below.
Lawson and Hogben’s (1996) think-aloud study
Lawson and Hogben (1996) observed the behaviour of 15 university students in Australia, with experience in Italian, as they attempted to learn the meanings of new Italian words. The researchers employed a think-aloud procedure, and a vocabulary post-test to examine the efficiency of the strategies selected by the learners during an intentional L2 vocabulary learning task. The participants were asked to attempt to learn new L2 target words presented on index cards. These were accompanied by examples of sentences composed of the target words that appeared on one side of the card, with translations and related words on the other. During the study phase, the participants were asked to report on their thoughts and discussions while attempting to learn the target words. The majority of the procedures employed by the participants involved 15 categories of strategies that were grouped into four higher-level categories: repetition, word feature analysis, simple elaboration, and complex elaboration. Lawson and Hogben (1996) found that participants employed the following strategies, see Table 3.7.
121
Table 3.7: Strategies used in Lawson and Hogben’s (1996) study.
Category Strategy Frequency of use
Repetition
reading of related words 156
simple rehearsal 137
writing of word and meaning 45
cumulative rehearsal 15
testing 5
Word feature analysis spelling 16
word classification 9
Simple elaboration
sentence translation 66
simple use of context 36
appearance similarity 18
sound link 5
Complex elaboration
complex use of context 42
paraphrase 28
mnemonic 7
The researchers also found that the more often a strategy was used, the more it contributed to word learning. A strong positive correlation between the frequency of strategy use and the target word recall was revealed, as well as a positive correlation among the target word recall and simple rehearsal, mnemonic, appearance similarity, sound link, and paraphrase.
Barcroft’s (2009) self-reporting study
Barcroft’s (2009) study expanded upon that of Lawson and Hogben (1996) by examining the relationship between strategy use and vocabulary learning performance during intentional word-picture vocabulary learning, with the inclusion of new methodological provisions. The research was conducted with 93 English-speaking Spanish students as they studied new Spanish words, while viewing word-picture pairs. Barcroft selected the word- picture vocabulary learning paradigm “to focus on the strategies and cognitive activity in which learners engage when they are given access to each target word form and its referent” (Barcroft, 2009, p. 78). The participants answered questions concerning the strategies they employed. Two kinds of word recall test followed the learning phase: (1) a picture L2 recall test, which was a productive test that required the participants to write the target Spanish words when presented with pictures; and (2) a L2-L1 recall test, which was a receptive test that required the
122
participants to translate the target Spanish words when presented only with those words. All of the strategies were coded according to the participants’ written responses to questions one and two, after which the author and an assistant collaborated on the strategy coding.
The results of Barcroft’s study reported on the 12 most frequently-used strategies, which included L2-picture association, L2–L1 association, L2–L1 translation, and repetition. He found that some strategies, such as mnemonic, self-generated target words, and visualized target words and pictures resulted in a higher target word recall than others. Similar findings were discovered in Lawson and Hogben’s (1996) study concerning significant positive correlations between the numbers of strategies that learners reported using, and their target word recall. However, neither study examined the effects of intentional or incidental learning strategies on long-term word retention. Therefore, the present study examined the potential effects of VLS in terms of both intentional and incidental learning, on both immediate- and delayed word retention, and identified the relationship between these strategies, and vocabulary learning performance. The examination of the long-term effects enabled the categorization of the most frequently-used learning strategies that influence the effectiveness of L2 vocabulary presentation and practice methods.
3.5.3 Studies concerning incidental vocabulary learning
This section concerns a review of certain key studies regarding incidental vocabulary acquisition from reading, due to their relevance to the present study.
In the field of second and foreign language vocabulary acquisition, many studies have investigated the incidental acquisition of vocabulary through reading (e.g., Brown et al., 2008; Horst, 2005; Hulstijn et al., 1996; Kweon & Kim, 2008; Pellicer-Sánchez & Schmitt, 2010; Pigada & Schmitt, 2006; Rott, 1999; Vidal, 2011; Waring & Takaki, 2003). As previously discussed, this mode of learning is often referred to learning from context, since the meaning of the new lexical items should be inferred from the cues, and the meaning of the surrounding
123
words, and from the context of the sentence. As is evident in this review, these studies explored the effect of the number of exposures to the new words, which varied from one exposure to more than 20 exposures. For example, Hulstijn et al. (1996) examined incidental vocabulary learning under three conditions: (1) marginal glosses; (2) dictionary use; and (3) reoccurrence of unknown words in the reading text. The authors divided the participants, who were 78 advanced Dutch learners of French, into three groups: (1) Marginal glosses; (2) Dictionary; and (3) Control. The participants in the marginal glosses group received L1 translations of the new target words, those in the dictionary group were allowed to use a dictionary while reading, while those in the control group received neither a dictionary, nor marginal glosses. The reading text provided to the participants was a short story, the text of which was modified for the marginal glosses group by adding marginal glosses for the 16 target words in the right-hand margin, along with glosses for 16 additional words. According to the researchers, the additional 16 words were included “to render the marginal glosses for the 16 targeted words less salient” (Hulstijn et al., 1996, p. 330).
In the learning phase, the participants were instructed to spend 25 minutes reading the text. In order to meet the conditions of incidental vocabulary learning, the participants were told in advance that they would be tested only on their comprehension of the text, and not on their knowledge of unfamiliar words. Accordingly, the participants focused on the meaning of the entire text, rather than on the new vocabulary. However, the participants were later unexpectedly tested on these unknown words. Following the learning phase, Hulstijn et al. (1996) administered three post-tests, the first of which was a recognition and recall test containing a list of 32 words, which included the 16 target words, and 16 words from outside the reading text. The second test examined the students’ pre-knowledge of the target words, while the third test assessed the target words in their contexts. The findings revealed that the frequency of target words present in the reading text, from one to three times, significantly
124
supported incidental learning in all three groups: marginal glosses, dictionary, and control. The study also found that the effect of marginal glosses was greater than that of dictionary use, primarily because the learners did not consult their dictionaries.
However, one limitation of the study was that it did not examine the learners’ pre- knowledge of the target words before conducting the learning phase through reading, which may have engendered some previous knowledge of the target words. Thus, to avoid this limitation, the present study tested learners’ previous knowledge of the target words before conducting the incidental learning session to ensure that the participants possessed no prior knowledge of the target words. Moreover, Hulstijn et al. (1996) did not investigate long-term uptake knowledge in their study, or the vocabulary learning strategies employed during incidental learning, issues that were investigated by the current study
Meanwhile, Rott (1999) conducted a study involving 67 intermediate German language learners studying at a university in the USA. In preparing the reading materials, Rott composed a set of short paragraphs of four to six sentences each, involving new words that the L2 learners were to read once per week. Using a between-subject design, the subjects were assigned to two treatment groups, and each group was exposed to one set of the target words, with six target words in each set. Each treatment group was further divided into three sub-groups, to examine the effect of exposure frequency on the acquisition of the target words in terms of either two, four, or six times. Before commencing the reading session, Rott ran a pre-test to ensure that the target words were unfamiliar to the participants, employing a checklist format containing 50 words, including the 12 target words, along with 38 distracters. The subjects were asked to provide the meaning of the words they knew, and to skip those that were unknown to them.
Aiming to assess the participants’ knowledge of the target words, Rott employed two post-test formats, a productive vocabulary test of supply-definition, and a receptive vocabulary test of select a definition. The findings obtained from this study revealed that two exposures
125
could engender significant target word acquisition, since the scores for the items from two exposures were significantly higher than the scores for a control set of items to which the participants had not been exposed. However, the participants who were exposed to target words six times outperformed the participants who had only been exposed to target words two or four times. In addition, Rott’s research indicated that incidental learning can occur through a few exposures to new words in meaningful context. This study holds significance for the present study, since similar procedures to that used by Rott, in terms of the preparation of the reading materials, and conducting a productive vocabulary test, were employed.
Meanwhile, Waring and Takaki (2003) conducted a seminal study concerning incidental L2 vocabulary acquisition by examining the vocabulary learning and retention of 15 intermediate Japanese college learners of English. The participants were asked to read and enjoy a graded book, which was at a lower level than their language ability. According to Waring and Takaki (2003), the motivation behind providing the learners with an easy text was to ensure that the adjacent words were familiar enough to assist in inferring the meaning of the target words. Aiming to govern the learners’ previous knowledge of the target words, the researchers changed the forms of the 25 words that were used in the text into non-words, for example, changing house into windle, and window into bettle). The target words were varied in terms of frequency of occurrence in the text from one to 18 occurrences.
Unlike other studies that employed only one type of test, for example, a multiple-choice test, Waring and Takaki used three different types of measurements: a simple yes or no sight- recognition test, a standard multiple-choice test, and a meaning translation test into the first language. Their study investigated the effect of frequent exposure to new words in a reading text, on the words’ acquisition and retention over time, and was the first study to investigate word retention of incidental vocabulary acquisition over time (Alahirsh, 2014). The three tests
126
were administered three times over the course of three months: (1) immediately after the participants had finished reading; (2) after one week; and (3) after three months.
The results demonstrated that incidental vocabulary learning from reading occurred at several levels, and the scores obtained varied between the test types. From the 25 target words, there was a gain score of 15.3 words on the form recognition test, while in terms of meaning recognition via a multiple-choice test, the mean score was 10.6. However, the meaning translation test yielded a lower mean score of 4.6. The results obtained from the post and delayed tests showed that approximately half of the word knowledge gained from the reading was lost after three months. Furthermore, the study indicated that there was a robust relationship between word frequency in a text, and its retention. However, Waring and Takaki concluded that even with the words that occurred more than 18 times, the chance of retaining their meaning after three months was only between 10 and 15%. Additionally, the results also showed that there was no instance of words being retained if they were encountered fewer than eight times, although this result in particular conflicted with findings of Rott (1999), and Hulstijn et al. (1996), in the way in which incidental learning can occur through a limited number of exposures of one to three times in meaningful context. However, Waring and Takaki’s findings were supported by those of other studies, such as Brown et al. (2008) and Pigada and Schmitt (2006).
Waring and Takaki (2003) is of significance to the present study in the way that they employed more rigorous methodological procedures, as the authors used more sensitive tests, by running three vocabulary testing formats over the course of three months. The word meaning translation test appeared to be the most demanding test, since it provided more information regarding whether the answers in the multiple-choice test were the result of the subjects making inferences, or of actual learning of the assessed words (Alahirsh, 2014). The findings of this study encouraged the researcher of the present study to adopt a word meaning
127
translation test as the main test for this research, making it possible to measure the learners’ productive vocabulary knowledge resulting from employing different types of strategies more precisely.
Waring and Takaki’s (2003) study also suffered from certain limitations. For example, while they examined word retention of incidental learning over time, they did not examine the strategies behind this type of learning mode, focusing only on the effect of the number of exposures to the target forms. Furthermore, the study also failed to investigate the effect of other aspects of word knowledge in incidental learning, such as the words’ type of speech, in terms of noun versus verbs, and the effect of concreteness and abstractness on word retention within the incidental mode of learning. Another issue requiring consideration is the use of non- words as experimental target words, and the question of whether it is ethically appropriate to require participants to learn such words, in terms of whether the participants were aware of the fact that they were non-words, and whether this affected their incidental learning in some way. In a further study, Pigada and Schmitt (2006) examined the incidental learning involved in extensive reading in terms of different aspects of word knowledge: meaning, in the sense of form meaning relationship; form, in terms of spelling; and grammatical behaviour, or use. This research involved a case study with a Greek native speaker studying French, who was given one month of extensive reading before measuring his incidental acquisition of the target words. For the reading materials, the ‘Lectures CLE en Français facile’ graded readers (Level 1) were employed, according to the learner’s lower intermediate level in French language. Then, to examine his knowledge of 133 target words, a one-to-one interview with the participant was undertaken before and after the reading. The target words consisted of 70 nouns and 63 verbs, and were divided into six frequency groups of one, two to three, four to five, six to 10, 10 plus, and 20 plus occurrences). The participant was not aware of the aim of the study, or of the post- test that occurred after the reading session. There were two test types, a pre- and post-test,
128
which were taken over two days, a spelling test on the first day, and meaning and grammatical use on the second day. The subject was given an irrelevant task to perform after completing the spelling test, which “served to flush the target words from [the subject’s] immediate memory” (2006, p. 11), so that the researchers could guarantee that no additional intentional learning of these words would occur after the session. Additionally, the researchers changed the word order in the meaning and grammatical use tests, to avoid parallelisms in the test form.
The findings revealed that the participant displayed a 65% improvement in at least one of the knowledge aspects of the target words, and the participant’s spelling knowledge improved meaningfully, from 37% to 60% in the post-test. Another interesting finding was that the participant’s knowledge of nouns was better enhanced for all six frequency groups in comparison with their knowledge of verbs, which showed no increased in the six to 10, and 20 plus verb occurrences groups. In general, the participant’s knowledge of words’ meaning was improved considerably in all frequency groups, however, only small amount of learning was reported for nouns occurring once or two to three times. Similarly, there was no learning of verbs that were encountered only once in the texts. These results concurred with Waring and Takaki’s (2003) findings, but contrasted with those of Rott (1999), in that incidental learning can occur through a small number of exposures to new words in a meaningful context. Nevertheless, it is worth noting that Rott composed the reading material specifically for the study, and hence each encounter with the target words would likely have been in a very clear context, which may have facilitated incidental acquisition. Overall, Pigada and Schmitt (2006) results demonstrated that the understanding of approximately half of the target words (49.6%) improved in one knowledge aspect, 10% in two knowledge types, and 7% in all three knowledge aspects. Although the studies of Pigada and Schmitt (2006) and Rott (1999) confirmed the occurrence of incidental learning through reading in long and short texts
129
respectively, neither investigated the reasons or strategies that cause such a learning to occur, nor did they investigate delayed retention, a matter that the present study has addressed.
In a more recent study of incidental vocabulary learning, Pellicer-Sánchez and Schmitt (2010) required their participants to read a 150 page novel, including 34 unknown target words. The target words were divided into 5 frequency bands, with the exposure to the words in the story altering from one to 28 occurrences. When the participants had completed the novel, the researchers employed four types of tests to measure their vocabulary knowledge: spelling recognition, word class knowledge, meaning recognition, and meaning recall tests. The results revealed that knowledge gains were found in nine of the target words, and it also revealed the effect of frequency of exposure, in terms of the fact that the learners had significantly more knowledge across all the tests regarding the words that appeared 10 or more times in the novel, compared with the words that appeared eight times, or fewer. This result concurred with the